To define the direct toxicities of antiretroviral drugs (ARV) and separate those changes from their effects on HIV infection, we gave single ARV to healthy volunteers and compared results to studies in HIV infected patients on those ARV. We now propose to resolve controversies with currently used ARV. 1) The mechanism(s) by which ritonavir increases triglycerides is unknown, but have significant implications for atherosclerosis. Hypothesis 1: Ritonavir-based regimens increase triglycerides and VLDL by both increasing VLDL production and decreasing VLDL clearance. Specific Aim 1A: To quantify the effect of ritonavir on VLDL production and clearance using stable isotope turnover and other clearance methods. Specific Aim 2B: To determine the composition of the triglyceride rich particles 2) NNRTI increase HDL cholesterol, but the mechanism is unknown and NNRTI may not generate HDL particles that have strong anti-atherogenic effects. Hypothesis 2: NNRTI drugs do not increase HDL by increasing apo AI production, but by decreasing apo AI clearance, prolonging time in circulation and producing particles with limited anti-atherogenic properties. Specific Aim 2A: To determine the composition of HDL before and after NNRTI and assess its function. Specific Aim 2B: To quantify the effect of NNRTI on apo AI production & clearance using stable isotopes. Specific Aim 2C: To determine if the efavirenz induced increase in HDL is accompanied by improvement in flow mediated vasodilation and circulating markers of endothelial function 3) The effects of PI on glucose metabolism cannot be solely explained by their effects on insulin resistance. PIs may impair insulin secretion, which has important implications for design of safer PI. Hypothesis 3: Ritonavir-based PIs impair insulin secretion. Specific Aim 3: To determine which ritonavir-based PI regimens alter insulin secretion. The results from these studies will define the direct toxicities of ARV, which provide essential information for rationally evaluating treatment strategies and counseling patients.Carl Grunfeld, MD, PhD. [unreadable] [unreadable] [unreadable]